Epilepsy affects an estimated 2.5 million people in the United States and is associated with a high risk of progressive cognitive and psychosocial dysfunction, and enormous socioeconomic and health-care utilization costs. There is currently little understanding of why some patients respond well to anticonvulsant therapy whereas others develop uncontrolled seizures and progressive brain dysfunction. Powerful imaging tools are now available for quantitatively characterizing the structural and functional connections between brain regions that make up epileptic networks, providing a promising new approach for understanding, predicting, and treating refractory epilepsy. Temporal lobe epilepsy (TLE) is the most common form of epilepsy in adults and the largest group among those with medically refractory seizures. The Epilepsy Connectome Project (ECP) will collect detailed connectivity measurements in 200 people with idiopathic TLE using diffusion-weighted magnetic resonance imaging of structural connections, functional magnetic resonance imaging of dynamic network interactions, and magnetoencephalography to measure these interactions with millisecond time resolution. The methods will closely mirror those currently used by the Human Connectome Project (HCP) to study network connectivity in healthy participants, and the HCP data will provide a critical baseline against which to compare the ECP connectome data. These comparisons, based on large cohorts studied with sensitive, state-of-the-art methods, will reveal for the first time the full extent of abnormal network structure and function in TLE. The data will be used to test four major hypotheses: 1) that recurring seizures over many years lead to connectivity abnormalities in TLE, 2) that connectivity abnormalities account for the cognitive and psychosocial dysfunction observed in people with TLE, 3) that the severity of connectivity abnormalities predicts the risk of subsequent decline in cognitive and psychosocial function, and 4) that the severity of connectivity abnormalities predicts the risk of developing medically refractory seizures. Evidence supporting these hypotheses would lead directly to novel clinical tools for diagnosis and individualized management of patients with epilepsy based on quantitative imaging of the connectome.